(WIP) New modchip coming in

Benny

Thanks for the reply and confirmation about the LCD header.

I do agree that on a new project like this and with the production run that you are proposing, keeping build quality/reliability up while keeping costs to a minimum have to be a priority. And as you rightly say - not everyone will want to hook up an lcd and even if they do, they can solder direct to the pads on the chip and those that want a connector can add one themselves. That seems pretty fair.

So far as pricing goes, what you are proposing seems absolutely remarkable value for all the features that are crammed into this little beauty - I imagine they are going to be pretty popular!

Cheers, Mark

 

Small update.

I've been hitting my head on a bug on the TSOP recovery feature the whole weekend(A15 signal line wouldn't work when using TSOP split feature). The good news is that I had an epiphany on my way to work this morning so there's a good chance it'll be fixed tonight! The bad news is that it pushes my schedule back about a week since I'm cautious and I now want to test the whole design again!

On the OS side, I worked on the 128MB RAM test code. RAM tester will test each extra RAM chips individually instead of testing bank by bank(which are distributed evenly across physical RAM chips and thus all 4 banks will report failure even if just 1 of the 4 chips is not properly installed). So I tested this myself during the weekend. I took a 1.0 and a 1.2 and ran RAM tests between each chip installation. Both motherboards maps their extra RAM chips the same way. Here's the numbering convention used in XBlast OS to identify RAM chips mapped to the upper 64MB:




This will greatly help spot faulty chips. I didn't want to specify which pin might cause trouble because I figure if at least one pin of the chip is not properly soldered, you should take the time to recheck the whole chip to be sure (i.e. I won't promote lazy workmanship!).

 

do you made the sst chip in a socket or hardleaded? socket was better to switch it when the chip was broken...

 

Benny - you don't do this update any justice by calling it "small"!

I simply LOVE your memory test feature - such an improvement that we can test individual chips now and identify which one is faulty/been poorly installed - thank you.

Well done on the TSOP recovery feature progress too mate. I am sure it must be frustrating to be set back but what is a week between friends? At the end of the day, you are wise to be cautious and to be honest, it is your attention to detail and rigorous thinking/testing that is turning this into such a stonking little project.

Cheers, Mark

 

SST flash chip is now soldered on the PCB. Again this is to reduce costs:
-The cost of the socket itself
-Added cost to PCB fabrication due to enlarged size
-Added cost due to extra parts for proper decoupling

SST49LF080A is rated for at least 100 000 cycles. Before you go through that...
OS will always be on board(unless you manually flash over it but that's not something you can do easily) so no worry to reflash the user flash banks.
OS flash/update procedure is secured by checksum(CRC32) calculation prior to reflashing. The 2 only probable causes that could "brick" the modchip is a faulty flash chip(can be tested before shipping) or power outage during reflash. Even if the latter was to happen, you can recover by directly booting 1 of the user banks by jumping either "SW1" or "SW2" to ground on the modchip. Once the console is fully boot, remove jumper and flash XBlast OS update using either Evolution-X, XBlast OS(xbe) or any other supported flash tools. Just make sure to have at least 1 good BIOS on your banks before attempting to update OS.

Thanks for the words of encouragements. I must admit that this whole thing took quite some time. Disassembling console, soldering 1 RAM chip, cleaning and checking, reassembling and testing done 4 times simultaneously on 2 consoles is time consuming. But I did it right and I did it once so I can move on to something else!

It looks like I fixed the problem of TSOP recovery. Will test further to see if everything is holding up just right!


With the new RAM testing code(and all the changes since v0.1beta), I feel another beta release could be useful to some. I will make sure XBlast OS is still fully stable and working on other modchips and release another package. Most of my work focus on making this thing work on a XBlast Mod and so it makes calls to specific LPC registers for many features. I just need to make sure that those calls can be bypassed when no XBlast Mod is detected.

 

Put me down for the first batch please.

 

Would love one to review too

 

Haha, you can definitely expect that I'm going to be grabbing a few of these myself. Can't wait

 

I would love to get my hands on one from the first batch as well.

 

me too me too..

 

Depending on the cost, I might want to have one to help me learn how to mod xboxes

 

I made a new OS release.
It's in it own separate thread. I figure this might be better as feature-list grows, so does the probability of questions and help request. I don't want to overflow this thread with questions relating uses of XBlast OS without a XBlast Mod.

There is located over there:
http://www.assemblergames.com/forums/showthread.php?55028-Release-XBlast-OS-v0-2Beta

 

Due to the server disaster, all the posts the past month or so have been lost. Did you made any progress you'd like to share with us in January?
I have a softmoded xbox the past ten years directly hooked on my TV, and I'd like to get your excellent modchip too!

 

Robo

Your previous post in this thread (which was lost) was:

***************
Please forgive my ignorance if my question has already been answered, I wanted to ask if there's a planned release date, such as first quarter of 2015 or is it going to be much later?
In any case, you should consider making/ordering a large batch, there is still a demand of quality xbox modchips and I do believe yours are going to be an instant success.
***************

Benny answered with the following:

***************
I originally planned for a Q1 2015 release but I fear I will have to re evaluate that speculation.

Beta units are currently shipping to a few selected testers. I would like for all these testers to have a full month with the chip to make sure it is stable and fully functionnal. I'm confident that nothing will go wrong but it's a necessary step I will not skip.

That will bring us around mid march. Then we can start talking mass production, which will also take around a full month. So I'd say Q2 2015 looks more realistic now. Still it's pretty good considering I started working modchip code on Aladdin XTs around April-May of last year. A full working modchip with capable OS from a single developer under a year is good by me.

Anyway hang in there. A couple of months and it will be ready!
*****************

Hope that helps.

 

Kudos clabs! However there were several other good posts regarding the subject, I hope bennydiamond could shed some light regarding the latest development.

 

No worries Robo

They were the only 2 posts I had as email updates.

There were other posts lost but I think that one I quoted from Benny was one of the last ones and hopefully gets you up to speed.

In short - this will kick ass so hold fire until they are released - you wont be disappointed )

 

Hi there,

I have made some progress yes. I'm always working toward making this a great modchip!
Some Beta units have been sent to selected candidates, they should get them in max 2 weeks. I just want for them to install it and play around with it. I'm constantly doing tests of my own but there's nothing like a good old field test to bring out corner cases I wouldn't have thought about.

For now, I have split my time on 4 tasks:

1. Continuous improvement of XBlast OS
2. Creation of an atomated test platform for the PCB assembly factory to test units.
3. Creation of a website (wordpress, it's easy) to centralize information
4. Writing documentation for this thing (install manual, OS manual)

I'm gearing up for a 3rd release of the OS really soon. Main new features are:

• -Restore EEPROM from .bin backup file on Xbox HDD (in a fixed location for now)
• -Restore EEPROM from .bin backup file uploaded through the Web server (like the BIOS netFlash feature already present)
• -Script Engine. Only useful if you have the modchip for now. It allows you to write scripts and execute them, before booting a BIOS bank, to do all sorts of things.
  [*=1]Toggle LCD power
  [*=1]Display text strings on LCD
  [*=1]Change Fan speed
  [*=1]Change front panel LED color pattern
  [*=1]Read GPI(General Purpose Input) pinson the XBlast Lite
  [*=1]Toggle GPO (General Purpose Output) pins on the XBlast Lite
  [*=1]You can create numeric variables and do arithmetic operations
  [*=1]Engine can take input parameters when it is called. It stores them into variables and can be used in script.

here's a test script I made. It's complex for nothing but it test the engine pretty well.
[HR][/HR]

[HR][/HR]There are limitation to this engine. You cannot do a function call inside a function call (like "GPO(GPI(1), 0)").
You also cannot initialise a freshly created variable from a function call(like "VAR foo = GPI(1)", but "foo = GPI(1)" is acceptable if "VAR foo" was declared earlier)
There's no "ELSIF" statement

I'm sorry if syntax and coding style is not to your liking but I made this all up from scratch. There are scripting engines available out there without license restriction but they are too big and complex to my liking. Space is limited in this 256KB image file so I figured the best thing to do was to code exactly what I need instead of taking something really massive and try to trim it down.

Of course more testing will be needed to ensure it works properly but it seems to hold up pretty well with all the test scripts I throw at it.

The script engine was put in place partly to satisfy the request of fxmech who would like to toggle GPOs to switch HDD. It'll now be possible for him to implement such feature either from BIOS bank selection or physical switches connected to the GPI pins of the XBlast Lite.
There are also other applications to this. For example, you could mod your Xbox with some extra LEDs and make a script that would light up different patterns depending on which BIOS bank you boot.
​

So that's about it.

 

I wish I had seen this sooner. I'll have to take my time to comb through the tons of information here, but from what I've briefly seen, this is a pretty impressive modchip, and I for one would love one (well, one for NOW. In the future, I'll be after more).

Looks like you already have testers (and probably ones that are much more useful than I), but if you need another, I will happily sacrifice my brother's Xbox for you (it's a v1.6). =3

 

Another nice update Benny.

Sounds like you have been pretty busy on this and not leaving anything to chance. I am really looking forward to seeing the install and OS manual.

Keep it up mate!